Microcomputer Automated System for Measuring Systolic Time Intervals in Response to Exercise and a Psychophysiological Task

A microcomputer automated system for measuring systolic time intervals is described. Electrocardiogram, phonocardiogram, and carotid pulse tracings were measured in 38 healthy male subjects during baseline conditions and during either exercise on a bicycle ergometer or a video-game task. These measurements were recorded on both a traditional 3-channel ECG recorder and the computerized system. Both methods of recording systolic time intervals were independently scored by two different experimenters. In this way, interrater reliability of hand-scoring, intermethod reliability between hand-scoring versus computer-scoring, and interrater reliability of computer-scoring could be assessed. The interrater reliabilities of hand-scored systolic time intervals were generally above .90, ranging from .73 for left ventricular ejection time to .99 for R-R intervals of the ECG, with a mean of .92. The intermethod reliability of the computer versus hand-scored systolic time intervals also proved to be generally above .90, ranging from .76 for S1-S2 components of the phonocardiogram to .99 for R-R, with a mean of .94. The interrater reliabilities of the computer-scored systolic time intervals were all above .90, ranging from .93 for S1-S2 to .99 for R-R, with a mean of .98. These data indicate that the computerized method of scoring systolic time intervals is at least as reliable as the more traditional scoring of paper tracing.